silk.text.byte_search.

When to use#

Use these functions instead of a byte-by-byte loop when scanning text or binary data, for example to find the end of a line, a delimiter or a header name.

Details#

On CPUs with 128-bit vectors (x86-64, AArch64, and wasm32 with simd128), each function compares sixteen bytes at a time with silk.simd.u8x16 and finishes the last bytes one by one. Elsewhere it is a plain loop. The results are the same everywhere; only the speed differs. findSlice filters candidates by the first and last byte of the needle, then compares the whole needle at each candidate.

Examples#

Split off the first line#

import silk.text.byte_search
import silk.core.option { Option }

pub fn main() -> i32 {
  let text = b"GET / HTTP/1.1\r\nHost: example.com\r\n"
  let end = match byte_search.findSlice(text, b"\r\n") {
    Option<usize>.Some { value } => value
    Option<usize>.None => { return 1 }
  }
  if end != 14 || byte_search.count(text, 10) != 2 {
    return 1
  }
  return 0
}

Import as byte_search with import silk.text.byte_search as byte_search.

Public declarations: 6.

find#

pub fn find(haystack: &[u8], wanted: u8) -> Option<usize>

Returns the index of the first byte of haystack equal to wanted, or none.

findLast#

pub fn findLast(haystack: &[u8], wanted: u8) -> Option<usize>

Returns the index of the last byte of haystack equal to wanted, or none.

find2#

pub fn find2(haystack: &[u8], first: u8, second: u8) -> Option<usize>

Returns the index of the first byte of haystack equal to first or second, or none.

find3#

pub fn find3(haystack: &[u8], first: u8, second: u8, third: u8) -> Option<usize>

Returns the index of the first byte of haystack equal to first, second or third, or none.

count#

pub fn count(haystack: &[u8], wanted: u8) -> usize

Returns the number of bytes of haystack equal to wanted.

findSlice#

pub fn findSlice(haystack: &[u8], needle: &[u8]) -> Option<usize>

Returns the index where needle first occurs in haystack, or none. An empty needle occurs at 0.

Category: Text.

Profiles: aarch64-apple-darwin, aarch64-unknown-linux-gnu, aarch64-unknown-linux-gnu-no-libc, x86_64-unknown-linux-gnu, x86_64-unknown-linux-gnu-no-libc, wasm32-unknown-unknown, wasm32-wasip1, wasm32-wasip2.